Introduction

Consequences of stress generate major health concerns for young adults enrolled in higher education. In addition to contributing to numerous physical and psychological health problems (Donatelle, 2014; Karren, Smith, & Gordon, 2014), stress causes burnout (Deckro et al., 2002; Jenkins & Elliot, 2004; Karren et al., 2014; Pines & Keinan, 2005). Burnout can lead people to question their vocational choice and consider leaving their line of work (Pines & Keinan, 2005). Since many graduate students view their educational pursuits as a full-time job, stress that leads to burnout may increase attrition rates among this population. Both social support (Donatelle, 2014; Karren et al., 2014; Jenkins & Elliot, 2004; Bolt, 2004) and a sense of purpose in life (Donatelle, 2014; Jaret, 2016; Karren et al., 2014; Schaefer et al., 2013) provide a buffer against stress by helping individuals cope more effectively.

While universities have been urged to incorporate a holistic approach in addressing student needs, attempts to assist students in developing social support are often inadequate (Astin, Astin, Lindholm, & Bryant 2005; Christie, Munro & Fisher, 2004; Ellis, 2001; Williams, 2002) and the appeal to assist students in discovering a sense of meaning and purpose in life is often ignored (Astin, Astin, & Lindholm, 2003; Astin, Astin, & Lindholm, 2011; Astin et al., 2005; Dalton, 2001; Laurence, 2005; Love, 2001). Neglecting these important functions means young adults may not garner benefit from their stress-buffering effects.

While health benefits of social support and a sense of meaning and purpose in life have been investigated independently, few research studies have examined these variables with regard to the graduate student population and no research published to date has examined the interaction between these variables with regard to stress. This study sought to explore the association between stress and social support, life purpose, and selected demographic characteristics among graduate students, to explain how the main effects, as well as the interaction effect, of social support and life purpose influence stress levels of graduate students, and to determine if the influence of the interaction varies by selected demographic characteristics.

Method

Procedures & Participants

Using a cross-sectional, web-based survey research design approved by the Institutional Review Board (IRB) at a large southeastern university, surveys were distributed, completed, and returned via the internet. The web-based survey process was facilitated by SurveyMonkey, a service that provided a password-protected account from which the surveys were created and disseminated as well as from which survey responses were collected and data compiled. A random sample of 2,000 graduate students was generated by a SAS random numbers program from among 6,545 graduate students enrolled at a large southeastern university. These 2,000 graduate students were sent a pre-notification letter via e-mail informing them that they would receive an e-mail requesting their participation in a web-based survey. Twenty-one of the 2,000 pre-notification emails initially sent were returned as "undeliverable". Therefore, 1,979 graduate students received an e-mailed cover letter explaining the purpose of the study, inviting them to participate, and linking those who chose to participate to the web-based survey. Completion of the 78-item survey, which included demographic questions as well as measures of social support, life purpose, and stress, served as implied consent.

Instruments

Demographic variables were measured by asking participants to report their sex, age, race, field of study (college in which the graduate program departments of the participants were housed), type of degree (masters or doctoral), number of credit hours enrolled during the current semester, and time in program (total number of semesters enrolled in their current graduate program, including the current semester) as well as program focus (coursework, comprehensive/qualifying exams,

or thesis/ dissertation research). The variables social support, life purpose, and stress were measured utilizing instruments with demonstrated internal reliability, convergent validity, and discriminant validity. Social support was measured using a composite score of items within the Interpersonal Support Evaluation List (ISEL), a 40-item scale that assesses the perceived availability of potential social resources (Cohen, Mermelstein, Kamarck, & Hoberman, 1985). Life purpose was determined by the Personal Meaning Index (PMI), a 16-item scale focusing on existential beliefs that life is meaningful (Reker, 1992), which is a composite score of the Purpose and Coherence subscales of the Life Attitude Profile-Revised (Reker, 2005). Stress was assessed using a composite score of the items within the Perceived Stress Scale (PSS), a 14-item scale assessing perception of situations as stressful (Cohen, Kamarck, & Mermelstein, 1983).

Results

Participants

Of the 1,979 graduate students who received the email requesting their participation, 572 participants completed all of the survey items, yielding a 29% response rate. A majority of the participants were white (n = 475; 70.2%), female (n = 383; 56.3%), pursuing a master's degree (n = 382; 56.3%), and devoting a majority of time to coursework (n = 440; 64.9%). In terms of field of study, several colleges were represented by more than 10% of respondents: Liberal Arts and Sciences (n = 150; 22.1%), health-related colleges (Dentistry, Health and Human Performance, Medicine, Nursing, Pharmacy, Public Health and Health Professions) (n = 112; 15.68%), Business Administration (n = 100; 14.7%), Education (n = 93; 13.7%), Agricultural & Life Sciences (n = 90; 13.2%), and Engineering (n = 82; 12.1%). Most participants were between 21 and 30 years of age (n = 490; 72.8%), with a mean age of 28.6 (SD = 6.6). Participants were currently enrolled in an average of 9.7 credit hours (SD = 2.6), with a range of 2 to 18, and therefore most were full-time students (n = 481; 71.6%). The number of semesters enrolled in their current graduate program ranged from 1 to 30, with 5.2 semesters being the average (SD = 3.8).

Social Support, Life Purpose, & Stress

Composite scores for social support, life purpose, and stress were calculated as were ranges of values for these variables identified by determining the highest and lowest possible scores and labeling the top third of the scores as high, the middle third of the scores as moderate, and the bottom third of the scores as low. With possible scores on the ISEL ranging from 40 to 160, participants' social support scores fell within the upper third of possible scores, ranging from a low of 58 to a high of 160 (M = 130.6, SD = 17.3). With possible scores on the PMI ranging from 16 to 112, participants' life purpose scores fell within the upper third of possible scores, ranging from a low of 25 to a high of 112 (M = 81.4, SD = 17.4). With possible scores on the PSS ranging from 14 to 70, participants' stress scores fell within the middle third of possible scores, ranging from a low of 17 to a high of 67 (M = 39.3, SD = 7.8). Therefore, graduate students in this study experienced relatively high levels of both social support and life purpose as well as relatively moderate levels of stress.

Associations

Correlations, means, and standard deviations described the association between stress and selected demographic characteristics among graduate students. Correlations were calculated to determine the association between stress and social support, life purpose, and selected quantitative demographic characteristics. All significance tests were conducted at the .01 alpha level. Results indicated that, of the quantitative independent variables, stress was significantly correlated in a negative direction with both social support, r = -.393, p = .000, and life purpose, r = -.470, p = .000, among these graduate students. Table 1 displays the correlations between stress and all of the quantitative independent variables. Means and standard deviations for stress as a function of the categorical independent variables are presented in Table 2. The means ranged from 37.8 to 40.4, falling within the middle third of possible scores on the PSS.

A multiple linear regression analysis determined the strength of association between the dependent variable of stress and the independent variables of social support, life purpose, sex, age, race, field of study, type of degree, credit hours, time in program and program focus (see Table 3). All significance tests were conducted at the .01 alpha level. Results revealed a significant adjusted [R.sup.2] of .272, F (10, 540) = 21.51, p = .000, indicating that 27.2% of the total variance in the stress score was explained by social support, life purpose, sex, age, race, field of study, type of degree, credit hours, time in program and program focus.

Of these independent variables, social support, life purpose, and sex significantly contributed to the total variance in the stress score. Specifically, an analysis of the unstandardized regression coefficients of these variables revealed that for each unit increase in stress score, the level of social support of the participants decreased by .101 units (b = -.101, t (540) = -5.349, p = .000) and their sense of purpose in life decreased by .153 units (b = -.153, t (540) = -8.448, p = .000). In addition, being male was associated with lower stress scores (b = -3.136, t (540) = -5.588, p = .000). All significance tests were conducted at the .01 alpha level.

An analysis of the [R.sup.2] increase for social support and life purpose allowed for a determination of improvement in fit of the regression line when these predictor variables were taken into consideration. The increase in [R.sup.2] that occurred when social support was added to the regression equation ([DELTA] [R.sup.2] = .038), revealed that, of the 27.2% of total variance in stress score that was explained by the predictor set, 3.8% was uniquely associated with social support. The increase in [R.sup.2] that occurred when life purpose was added to the regression equation ([DELTA] [R.sup.2] = .095), revealed that, of the 27.2% of total variance in stress score that was explained by the predictor set, 9.5% was uniquely associated with life purpose.

Interactions

To determine whether the interaction between social support and life purpose influenced the stress levels of graduate students a multiple regression analysis was performed (see Table 4). All significance tests were conducted at the .01 alpha level. Results revealed a significant adjusted [R.sup.2] of .295, F (4, 566) = 60.752, p = .000, which indicated that 29.5% of the total variance in the stress score was explained by sex, social support, life purpose, and the interaction of these two variables (social support X life purpose). Sex was included in this multiple linear regression analysis based on the fact that it was one of the variables that significantly contributed to the total variance in the stress score. The results revealed that the interaction term was not statistically significant (p = .812). However, additional results from this multiple linear regression analysis did confirm the presence of significant main effects for social support (p = .000), life purpose (p = .000), and sex (p = .000).

Discussion

Social Support, Life Purpose, & Stress

The high levels of social support identified among graduate students in the current study differ from previous research findings, which report that graduate students tend to perceive the support they receive to be inadequate (Greene, 2015; Ellis, 2001). This discrepancy may be due to the fact that previous research explored social support available through institutions of higher education (i.e., faculty, advisors, peers), while this study asked participants to consider social support in general, which could come from a variety of relationships (i.e., faculty, advisors, peers as well as family, friends, co-workers).

The high levels of life purpose identified among graduate students in the current study also differ from previous research findings. For example, the Higher Education Research Institute study (Astin et al., 2003) found that as students accumulated more years of higher education, their level of life purpose declined. While previous studies explored levels of life purpose among undergraduate students, this study was unique in its focus on graduate students. Meaning and purpose in life guide young adults in selecting a vocation, so undergraduate students may feel their lives lack purpose because they have not yet selected a vocation. As young adults develop a sense of meaning and purpose in life, allowing their calling to guide their career path, they may ensure that their actions and beliefs match (Donatelle, 2014; Dalton, 2001). Thus, pursuing a graduate degree represents one way young adults aspire to fulfill an identified purpose (Strange, 2001). So, graduate students would be more likely to report a sense of purpose in their lives.

The moderate levels of stress identified among graduate students in the current study also differ from previous research, which concluded that stress constitutes a major health concern for young adults in higher education (Donatelle, 2014; Deckro et al., 2002). However, high levels of social support and life purpose, as reported by the subjects in this study, can provide a buffer against stress (Donatelle, 2014; Jenkins & Elliott, 2004; Bolt, 2004). Thus, stress levels among subjects in this study would be predictably lower than for individuals whose levels of social support and life purpose were not as high.

Associations

Findings from the current study revealed that social support, life purpose, and sex contributed significantly to total variance in the stress score. Independent of one another, both social support and life purpose were inversely related to stress (i.e., as social support increases, stress decreases and as life purpose increases, stress decreases). These findings are consistent with previous research, which indicates that both social support and life purpose prove independently beneficial in the lives of young adults because these factors provide a buffer against stress (Donatelle, 2014; Bolt, 2004; Clara, Cox, Enns, Murray, & Torgrudc, 2003; Hodges, 2002). Current findings also revealed that life purpose is better than social support at predicting stress levels. In addition, stress levels were found to be lower among male graduate students than among female graduate students.

Interactions

Findings from the current study also revealed the absence of a significant interaction effect between social support and life purpose, indicating no differential effects for social support across life purpose with regard to stress. Thus, contrary to expectations, graduate students with high social support and high life purpose did not report low stress levels, nor did those with low social support and low life purpose have high stress levels.

Recommendations

Social support and life purpose independently provide a buffer against stress (Donatelle, 2014; Bolt, 2004; Clara et al., 2003; Hodges, 2002). This stress buffering effect can benefit young adults by helping them cope more effectively, thereby reducing the likelihood they will experience negative physical and psychological health outcomes associated with stress (Donatelle, 2014; Deckro et al., 2002;). In addition, this stress buffering effect can also benefit young adults academically by decreasing the likelihood of burnout (Donatelle, 2014; Jenkins & Elliott, 2004, Bolt, 2004) and attrition (Pines & Keinan, 2005; Reed & Giacobbi, 2004). In order for these benefits to be realized within the realm of higher education, institutions would do well to reconsider how they attempt to provide social support to students (Christie et al., 2004; Ellis, 2001; Williams, 2002) as well as begin to offer opportunities for the discussion and development of life purpose among students (Astin et al., 2005; Dalton, 2001, Astin et al., 2003; Astin et al., 2011; Laurence, 2005; Love, 2001). Findings from the current study can inform professional practice by providing insight on the importance of incorporating social support and life purpose into the higher education experience.

University faculty, health care center staff and student affairs staff dedicate a considerable amount of time and resources implementing health promotion initiatives designed to reduce stress because of the negative physical and psychological health outcomes stress can elicit among young adults (Donatelle, 2014; Deckro et al., 2002). The current study revealed that students with higher levels of stress tend to have lower levels of social support and life purpose. Understanding this association will enable university faculty, health care center staff and student affairs staff to identify and advocate the need for health promotion initiatives designed to enhance social support and life purpose to buffer stress.

When implementing such initiatives, recognizing which students have higher levels of stress and lower levels of social support and life purpose can assist university faculty, health care center staff and student affairs staff in directing the initiatives toward those most in need, instead of attempting to reach the entire student body, many of whom may not need the services offered by such initiatives. Female graduate students, for example, were found to experience higher stress levels than male graduate students. Therefore, initiatives aimed at promoting social support and life purpose to buffer stress may produce a greater positive impact if targeted toward female graduate students. However, findings from the current study also revealed no interaction effect between social support and life purpose with regard to stress. This lack of interaction suggests that health promotion initiatives designed to reduce stress need not include both social support and life purpose, but would do better to focus on these issues individually. According to the current findings, focusing such health promotion initiatives on life purpose should be higher priority since it was found to be better than social support at predicting stress levels. As such, higher education institutions should be more intentional about offering opportunities for the discussion and development of life purpose (Astin et al., 2005; Dalton, 2001; Astin et al., 2003; Astin et al., 2011; Laurence, 2005; Love, 2001), both within and outside of the classroom. Putting these recommendations into practice would likely conserve scarce resources (i.e., time, money, personnel) and enhance positive outcomes (i.e., life purpose, social support, stress reduction).

References

Astin, A.W., Astin, H.S., & Lindholm, J.A. (2011). Cultivating the spirit: How college can enhance students inner lives. San Francisco, CA: Jossey-Bass.

Astin, A.W., Astin, H.S., & Lindholm, J.A. (2003). A summary of initial findings from pilot survey 2000-2003. Retrieved from Higher Education Research Institute website: http://spirituality. ucla.edu/docs/reports/A%20Summary%20of%20Initial%20 Findings%20(Survey%20Report).pdf

Astin, A.W., Astin, H.S., Lindholm, J.A., & Bryant, A.N. (2005). The spiritual lives of college students: A national study of students' search for meaning and purpose. Retrieved from Higher Education Research Institute website: http://spirituality.ucla.edu/docs/reports/ Spiritual_Life_College_Students_Full_Report.pdf

Bolt, M. (2004). Pursuing human strengths: A positive psychology guide. New York, NY: Worth Publishers.

Christie, H. Munro, M., & Fisher, T. (2004). Leaving university early: Exploring the differences between continuing and non-continuing students. Studies in Higher Education, 29(5), 617-636.

Clara, I.P., Cox, B.J., Enns, M.W., Murray, L.T., & Torgrudc, L.J. (2003). Confirmatory factor analysis of the Multidimensional Scale of Perceived Social Support in clinically distressed and student samples. Journal of Personality Assessment, 81(3), 265-270.

Cohen, S., Kamarck, T, & Mermelstein, R. (1983). A global measure of perceived stress. Journal of Health and Social Behavior, 24(4), 385-396.

Cohen, S., Mermelstein, R., Kamarck, T., & Hoberman, H.M. (1985). Measuring the functional components of social support. In I.G. Sarason, & B.R. Sarason (Eds.), Social support: Theory, research and applications (pp. 73-94). Boston, MA: Kluwer Boston, Inc.

Dalton, J.C. (2001). Career and calling: Finding a place for the spirit in work and community. New Directions for Student Services, 95, 17-25.

Deckro, G.R., Ballinger, K.M., Hoyt, M., Wilcher, M., Dusek, J., Myers, P.,.. .Benson, H. (2002). The evaluation of a mind/body intervention to reduce psychological distress and perceived stress in college students. Journal of American College Health, 50(6), 281-287.

Donatelle, R. J. (2014). Access to health. (13th ed.). Boston, MA: Pearson Education, Inc.

Ellis, E.M. (2001). The impact of race and gender on graduate school socialization, satisfaction with doctoral study, and commitment to degree completion. Western Journal of Black Studies, 25(1), 30-45.

Greene, M. (2015). Come hell or high water: Doctoral students' perceptions on support services and persistence. International Journal of Doctoral Studies, 10, 501-518. Retrieved from http://ijds.org/Volume10/IJDSv10p501-518Greene0597.pdf

Hodges, S. (2002). Mental health, depression, and dimensions of spirituality and religion. Journal of Adult Development, 9(2), 109-115.

Jaret, P. (2016). Does your life have purpose? Retrieved from University of California Berkeley Wellness website: http:// www.berkeleywellness.com/healthy-mind/mind-body/article/ does-your-life-have-purpose

Jenkins, R. & Elliot, P. (2004). Stressors, burnout and social support: Nurses in acute mental health settings. Journal of Advanced Nursing, 48(6), 622-631.

Karren, K.J., Smith, N.L., & Gordon, K.J. (2014). Mind body health: The effects of attitudes, emotions, and relationships. (5th ed.). Boston, MA: Pearson Education, Inc.

Laurence, P. (2005). Teaching, learning, & spirituality. Spirituality in Higher Education Newsletter, 2(2), 1-6.

Love, P.G. (2001). Spirituality and student development: Theoretical connections. New Directions for Student Services, 95, 7-16.

Pines, A.M. & Keinan, G. (2005). Stress and burnout: The significant difference. Personality and Individual Differences, 39, 625-635.

Reed, S. & Giacobbi, P.R. (2004). The stress and coping responses of certified graduate athletic training students. Journal of Athletic Training, 39(2), 193-200.

Reker, G.T. (1992). Manual for the Life Attitude Profile--Revised. Peterborough, ON: Student Psychologists Press.

Reker, G.T. (2005). Meaning in life of young, middle-aged, and older adults: Factorial validity, age, and gender invariance of the Personal Meaning Index (PMI). Personality and Individual Differences, 38, 71-85.

Schaefer, S. M., Morozink Boylan, J., van Reekum, C. M., Lapate, R. C., Norris, C. J., Ryff, C. D., & Davidson, R. J. (2013). Purpose in life predicts better emotional recovery from negative stimuli. PLoS ONE, 8(11), e80329. http://doi.org/10.1371/ journal.pone.0080329

Strange, C. C. (2001). Spiritual dimensions of graduate preparation in student affairs. New Directions for Student Services, 95, 57-67.

Williams, K. (2002). Minority and majority students' retrospective perceptions of social support in doctoral programs. Perceptual & Motor Skills, 95(1), 187-196.

Beth McKinney, PhD, MPH, CHES, Associate Professor of Health Promotion & Public Health, Lynchburg College

Table 1. Bivariate Correlation Coefficients Between Stress
and Quantitative Independent Variables

                                      Sociol     Life          Credit
Variables                    Stress   Support  Purpose   Age    Hours

Stress     Pearson
           Correlation
           Sig. (2-tailed)             .000     .0(5(0   .972   .322
           N                   654      593      624     648     645

Social     Pearson
Support    Correlation        -.393      1       .500   -.045   .015
           Sig. (2-tailed)    .000               .000    .266   .714
           N                   593      635      588     (205    602

Life       Pearson
Purpose    Correlation
           Sig. (2-tailed)    .000     .000              .773   .106
           N                   624      508      644     (735    633

Age        Pearson            - 001    - 045     011      1     - 243
           Correlation
           Sig. (2-tailed)    .972     .266      .773           .000
           N                   648     (505      (535    673     664

Credit     Pearson             039      015      064     243      1
Hours      Correlation
           Sig. (2-tailed)    .322     .714      .106    .000
           N                   645      602      633     664     671

Time in    Pearson
Program    Correlation
           Sig. (2-tailed)    .842     .147      .400    .000   .000
           N                  (550     (50(5     637     669     667

                             Time in
Variables                    Program

Stress     Pearson
           Correlation
           Sig. (2-tailed)     .842
           N                   650

Social     Pearson
Support    Correlation        -.059
           Sig. (2-tailed)     .047
           N                   606

Life       Pearson
Purpose    Correlation
           Sig. (2-tailed)     .400
           N                   637

Age        Pearson             147
           Correlation
           Sig. (2-tailed)     .000
           N                   669

Credit     Pearson             911
Hours      Correlation
           Sig. (2-tailed)     .000
           N                   667

Time in    Pearson
Program    Correlation
           Sig. (2-tailed)
           N                   676

Table 2. Levels of Stress as a Function of Categorical
Independent Variables

Variables             M      SD

Male                 37.8   7.6
Female               40.4   7.7
White                39.1   7.7
Non-White            39.9   7.7
Health-Related       39.6   7.7
Non-Health-Related   39.2   7.8
Masters              39.4   7.7
Doctoral             39.2   7.8
Coursework           39.1   7.7
Non-Coursework       39.6   7.9

Table 3. Multiple Linear Regression Unstandardized Regression
Coefficients, Standardized Regression Coefficients, t-Test Variables
Statistics, and R2 Increase for Study and Demographic

                            Std                               [DELTA]
Variables            b      Error   [beta]     t        p    [R.sup.2]

Intercept          64.492  3.041             21.208   .000
Social Support     -.101    .019    -.230    -5.349   .000      .038
Life Purpose       -.153    .018    -.358    -8.448   .000      .095
Sex                -3.136   .561    -.208    -5.588   .000
Age                -.004    .047    -.003     -.081   .936
Race                .152    .613     .009      .248   .804
Field of Study      .206    .181     .010      .262   .794
Type of Degree      .018    .638     .001      .028   .978
Credit Hours        .166    .111     .058     1.493   .136
Time in Program     .032    .098     .016      .330   .742
Program Focus      -.219    .728    -.014     -.301   .764


Table 4. Multiple Linear Regression Unstandardized Regression
Coefficients, Standardized Regression Coefficients and t-test
Statistics
                             Std
Variables             b      Error   [beta]      t       P

Intercept          40.696   .385             105.687   .000
Sex                -3.499   .548    -.227     -6.387   .000
Social Support     -.110    .018    -.245     -5.936   .000
Life Purpose       -.154    .018    -.353     -8.555   .000
Social Support      .000    .001     .009       .238   .812
  X Life Purpose